WO2025056629A1

WO2025056629A1 - Method for detecting markers of minimal residual disease

Info

Publication number: WO2025056629A1
Application number: PCT/EP2024/075406
Authority: WO
Inventors: Natasha HONORÉ; Jean-Pascal MACHIELS
Original assignee: Universite Catholique de Louvain UCL
Current assignee: Universite Catholique de Louvain UCL
Priority date: 2023-09-11
Filing date: 2024-09-11
Publication date: 2025-03-20
Anticipated expiration: 2026-03-11

Abstract

The present invention relates to a method for detecting the risk of disease relapse in a subject previously treated for said disease. In some embodiments, the method is cancer.

Description

METHOD FOR DETECTING MARKERS OF MINIMAL RESIDUAL DISEASE

FIELD OF INVENTION

[0001 ] The present invention relates to methods for predicting the risk of disease relapse, in particular the risk of cancer relapse.

BACKGROUND OF INVENTION

[0002] Squamous cell carcinoma of the head and neck (SCCHN) is the seventh most common malignancy and affects around 600,000 patients per year worldwide. Less than sixty percent of patients with locally advanced (LA) disease (Union for International Cancer Control stages III and IV) remain free of disease at three years despite aggressive multimodal local therapy with surgery and/or chemo-radiation.

[0003] Recent randomized trials that investigated treatment intensification in LA SCCHN failed to meet their primary endpoints. These results may partly be explained by the absence of markers able to identify minimal residual disease (MRD) after curativeintent therapy, and such markers could potentially improve patient selection. Studies conducted in different cancer types have demonstrated that circulating tumor DNA (ctDNA) has the potential to identify patients likely to recur with high specificity and sensitivity. The early detection of patients at high or low risk of cancer relapse after curative-intent treatment could dramatically modify patient surveillance and therapeutic strategies.

[0004] SCCHN is heterogenous with several disease locations. It has two main etiologies (tobacco/alcohol versus human papilloma virus (HPV) and each one has a different prognosis. In HPV-negative tumors, genomic alterations are mainly found in tumor suppressor genes. These factors make ctDNA detection challenging, and only limited data are available for SCCHN, particularly for HPV-negative SCCHN. Using a multianalyte digital polymerase chain reaction (PCR) assay, it was showed that HPV-16 ctDNA detection in plasma samples during post-treatment surveillance has high positive- predictive value (PPV) and a high negative-predictive value (NPV) for diagnosing disease recurrence, with a median lead time of 3.9 months in patients with HPV-driven oropharyngeal cancer (OPC). Another study used a deep- sequencing personalized assay (RaDaR™) to detect ctDNA in post-surgery samples in 17 HPV-negative SCCHN. In all patients who developed disease recurrence, ctDNA was detected prior to progression with lead times ranging from 3.6 to 8.4 months. To detect ctDNA, both approaches investigated targeted methodologies based on prior knowledge of the patient- specific tumor genomic landscape, or HPV presence.

[0005] To the best of the Inventors knowledge, no studies have shown that a tumoragnostic MRD assay can detect ctDNA with clinically meaningful specificity and sensitivity in unselected LA SCCHN. Such agnostic technologies have the advantage as they can detect ctDNA without the need for a tumor biopsy and do not require a costly and personalized assay for each patient. However, their sensitivity may be decreased given that several genes generally need to be sequenced, and deep sequencing does not have the same low limit of detection as PCR-based technologies.

[0006] In the present invention, the Inventors provide an agnostic MRD assay able to predict disease recurrence. Early MRD detection using this MRD assay is associated with a shorter progression-free survival (PFS) and overall survival (OS).

SUMMARY

[0007] This invention relates to a method for predicting a risk of cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(i) isolating cell free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said cfDNA molecules; (ii) sequencing the cfDNA molecules isolated, and optionally amplified, at step

(i);

(iii) based on the sequencing performed at step (ii), detecting, among the cfDNA sequences, the circulating tumor DNA (ctDNA) sequences isolated from said at least one biological sample collected before treatment and the ctDNA sequences isolated from the at least one biological sample collected after treatment; and

(iv) concluding that said subject is at risk of cancer relapse if at least one ctDNA sequence is detected at step (iii); wherein the cancer is squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is selected from the group consisting of: blood, plasma and serum.

[0008] This invention relates to a method for detecting a minimal residual disease (MRD) in a subject previously treated for a cancer, comprising the steps of:

(i) isolating cell free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said cfDNA molecules;

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i);

(iii) based on the sequencing performed at step (ii), detecting, among the cfDNA sequences, the circulating tumor DNA (ctDNA) sequences isolated from said at least one biological sample collected before treatment and the ctDNA sequences isolated from the at least one biological sample collected after treatment; and (iv) concluding that said subject has a MRD if at least one ctDNA sequence is detected at step (iii). wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is selected from the group consisting of: blood, plasma and serum.

[0009] This invention relates to a method for treating a cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(iii) based on the sequencing performed at step (ii), detecting, among the cfDNA sequences, the circulating tumor DNA (ctDNA) sequences isolated from said at least one biological sample collected before treatment and the ctDNA sequences isolated from the at least one biological sample collected after treatment;

(iv) concluding that said subject is at risk of cancer relapse if at least one ctDNA sequence is detected at step (iii); and

(v) administering to said subject a treatment for treating and/or preventing said disease if said subject is detected as being at risk of said disease relapse at step (iv) wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is selected from the group consisting of: blood, plasma and serum.

[0010] In some embodiments, the step (iii) further comprises comparing aligned sequence data obtained from germline DNA.

[0011] In some embodiments, the step (iii) comprises the following steps:

- performing sequence alignment on genome, preferably on human genome and optionally on virus-induced cancer genome;

- identifying variants of the cfDNA, such as single-nucleotide variants (SNV) and small indels variants;

- selecting the cfDNA variants, wherein the selected cfDNA variants meet one or more of the following criteria: (i) the cfDNA are not present in more than 1% across all healthy populations, according to the proportions reported in public databases such as Exac and Gnomad, (ii) the cfDNA variants are present in a least one read in both orientations, (iii) the cfDNA variants are on reads passing the quality filters from the GATK institution and Mutect2 pipeline, (iv) the cfDNA variants are not present in an intronic section of the genome, (v) the cfDNA variants are present on reads not exceeding 100 base pairs, (vi) the cfDNA variants are present in at least 3 reads, (vii) the cfDNA variants are at a position where at least 500 reads are aligned;

- measuring a cfDNA variant allele frequency (VAF) in the biological samples.

[0012] In some embodiments, at step (iii), a cfDNA variant is selected when: the cfDNA from said at least one biological sample collected before treatment is characterized in that: (a) at least three reads are detected for said cfDNA; (b) the reads are aligned in forward and reverse; and (c) the cfDNA is present at less than 1% of healthy individuals in public sequence databases; and/or the cfDNA from said at least one biological sample collected after treatment is characterized in that: (a) said cfDNA is also identified in the at least one biological sample collected before treatment; (b) at least three reads are detected for said cfDNA.

[0013] In some embodiments, a ctDNA molecule from at least one biological sample collected before treatment and a ctDNA molecule from at least one biological sample collected after treatment are considered identical if they share the same nucleic acid sequence.

[0014] In some embodiments, the at least one biological sample is plasma.

[0015] In some embodiments, the at least one biological sample collected after treatment is collected at least one week after the end of said treatment, preferably at least 12 weeks after the end of said treatment.

[0016] In some embodiments, more than one biological sample are collected after the treatment.

[0017] In some embodiments, the subject was treated for locally advanced squamous cell carcinoma of the head cancer or neck cancer (SCCHN); in particular stage III or stage IV locally advanced SCCHN.

[0018] In some embodiments, the ctDNA sequence comprises a nucleic acid sequence of any one of the genes selected in the group consisting of: AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof. [0019] In some embodiments, the ctDNA sequence comprises a nucleic acid sequence selected from E6 or E7.

[0020] In some embodiments, SCCHN is HPV-positive SCCHN, preferably HPV-16 positive SCCHN.

[0021] The invention also relates to a computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of the method as described herein.

[0022] The invention also relates to a computer-readable medium having stored thereon the computer program as described herein.

DEFINITIONS

[0023] In the present invention, the following terms have the following meanings:

[0024] “About”, when preceding a figure, means plus or less 10% of the value of said figure.

[0025] “And/Or” refers to and encompasses any and all possible combinations of one or more of the associated listed items, as well as the lack of combinations when interpreted in the alternative (“or”).

[0026] “At least one” includes 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 25, 50, 75, 100, 250, 500, 750, 10³,10⁴, 10⁵, 10⁶, 10⁷, 10⁸, 10⁹, 10¹⁰, 10¹¹, 10¹², 10¹³, 10¹⁴, 10¹⁵ or more.

[0027] “Comprising”, "comprises" and "comprised of" are used herein as synonymous with "including", "includes" or "containing", "contains", and are inclusive or open- ended and do not exclude additional, non-recited members, elements or method steps. These terms also encompass “consisting of’.

[0028] “Circulating DNA”, “cfDNA” or “cell-free DNA” herein refers to DNA that exists outside a cell in a subject or the isolated form of such DNA, typically in a body fluid, such as but not limited to, blood, plasma or serum. In some embodiments, the cfDNA molecule is a single-stranded or double-stranded DNA molecule, with a size a size ranging from 50 to 150 bp.

[0029] “Circulating tumor DNA’’ or ‘ ‘ctDNA” refers to a circulating DNA or cell-free DNA arising from a tumor cell. In some embodiments, the ctDNA molecule is a single- or double-stranded DNA molecule, with a size a size ranging from 50 to 150 bp. In some embodiments, ctDNA exists in plasma or serum. In some embodiments, ctDNA are released by the tumor cells into the blood and it thus harbors the mutations of the original tumor. Indeed, ctDNA possessed many cancer-associated molecular characteristics, including, without limitation, single-nucleotide mutations, methylation changes and cancer-derived viral sequences. In some embodiments, ctDNA are used as a liquid biopsy and significantly improve current systems of tumor diagnosis, even facilitating early- stage detection. In some embodiments, ctDNA is able to accurately determine the tumor progression, prognosis and assist in targeted therapy.

[0030] “Germline DNA”, refers to the genomic DNA of patient. The germline DNA is extracted from “normal” cells, i.e., non-cancerous cells. In some embodiments, the germline DNA is obtained from any tissue or fluid from the subject, preferably whole blood sample. In some embodiments, the germline DNA is subjected to amplification, sequencing and bioinformatic analyses according to the method of the invention. In some embodiments, the data recovered from the bioinformatic analyses according to the present invention is used as reference in variant calls analyses.

[0031] “Individual”, or “subject”, refers to an animal, preferably a mammal, more preferably a human. In one embodiment, the subject is a man. In another embodiment, the subject is a woman. In one embodiment, a subject may be a “patient”, i.e., a warmblooded animal, more preferably a human, who/which is awaiting the receipt of, or is receiving medical care or was/is/will be the object of a medical procedure, or is monitored for the development of a disease, preferably cancer. In one embodiment, the subject is an adult (for example a subject above the age of 18). In another embodiment, the subject is a child (for example a subject below the age of 18). [0032] “Minimal Residual Disease” refers to the disease that remains in a subject after treatment of a proliferative disease. Thus, determining the presence or absence of MRD means determining the presence or absence of diseased cells that remain proliferating in a subject or determining the presence or absence of genetic material that is associated with proliferative disease in a subject after treatment of said proliferative disease. Preferably, determining the presence or absence of MRD means determining the presence or absence of diseased cells that remain proliferating in a biological sample or tissue from a subject after treatment of said proliferative disease, or determining the presence or absence of MRD means determining the presence or absence of genetic material that is associated with proliferative disease in a biological sample or tissue from a subject, after treatment of said proliferative disease. The presence or absence of a proliferative disease may be identified based on the expression or lack of expression of a genetic marker on, in or outside diseased cells.

[0033] “Nucleic acid” or “polynucleotide” refers to any polyribonucleotide or polydeoxyribonucleotide, which may be unmodified RNA or DNA or modified RNA or DNA. “Nucleic acid” or “Polynucleotides” include, without limitation single-and doublestranded DNA, DNA that is a mixture of single- and double- stranded regions, single- and double- stranded RNA, and RNA that is a mixture of single- and double- stranded regions, hybrid molecules comprising DNA and RNA that may be single-stranded or, more typically, double-stranded or a mixture of single- and double- stranded regions. In addition, “Nucleic acid” or “polynucleotide” refers to triplestranded regions comprising RNA or DNA or both RNA and DNA. The term “nucleic acid” or “polynucleotide” also includes DNAs or RNAs containing one or more modified bases and DNAs or RNAs with backbones modified for stability or for other reasons. "Modified" bases include, for example, tritylated bases and unusual bases such as inosine. A variety of modifications has been made to DNA and RNA; thus, “nucleic acid” or “polynucleotide” embraces chemically, enzymatically or metabolically modified forms of polynucleotides as typically found in nature, as well as the chemical forms of DNA and RNA characteristic of viruses and cells. "Polynucleotide" also embraces relatively short polynucleotides, often referred to as oligonucleotides. [0034] “Protein”, “polypeptide” and “peptide” are used interchangeably herein to refer to polymers of amino acids of any length. The polymer may be linear or branched, it may comprise modified amino acids, and it may be interrupted by non-amino acids. The terms also encompass an amino acid polymer that has been modified by, for example, disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, pegylation, or any other manipulation, such as conjugation with a labelling component. As used herein the term “amino acid” includes natural and/or unnatural or synthetic amino acids, including glycine and both the D or L optical isomers, and amino acid analogs and peptidomimetic s .

[0035] “Sample” refers to any biological sample that is isolated from a subject. Examples of samples include, but are not limited to, an aliquot of body fluid, whole blood, platelets, serum, plasma, red blood cells, white blood cells or leucocytes, endothelial cells, tissue biopsies, synovial fluid, lymphatic fluid, ascites fluid, and interstitial or extracellular fluid. The term "sample" may also encompass the fluid in spaces between cells, including gingival crevicular fluid, bone marrow, cerebrospinal fluid (CSF), saliva, mucous, sputum, semen, sweat, urine, or any other bodily fluids. "Blood sample" can refer to whole blood or any fraction thereof, including blood cells, red blood cells, white blood cells or leucocytes, platelets, serum and plasma. The sample may be from a bodily fluid. The sample may be a plasma sample. The sample may be a serum sample. The sample may be a tumor sample. Samples can be obtained from a subject by means including but not limited to venipuncture, excretion, ejaculation, massage, biopsy, needle aspirate, lavage, scraping, surgical incision, or intervention or other means known in the art.

[0036] “Treating a disease” (in particular, “treating a cancer”) or “treatment” or “alleviation” refers to both therapeutic treatment and prophylactic or preventative measures; wherein the object is to prevent (i.e., keeping from happening) or slow down (lessen) a disease (such as, for example, a cancer) or an adverse effect or symptom thereof. Those in need of treatment include those already with cancer as well as those prone to have cancer or those in whom cancer is to be prevented. An individual or mammal is successfully “treated” for a disease, such as, for example, a cancer if, after receiving a therapeutic amount of a therapeutic agent, the individual shows observable and/or measurable reduction in or absence of one or more of the following: reduction in the number of aberrant cells or cancer cells; reduction in the percentage of total cells that are cancerous; and/or relief to some extent, one or more of the symptoms associated with the disease, such as, for example, the cancer; reduced morbidity and mortality, or improvement in quality of life issues. The above parameters for assessing successful treatment and improvement in the disease, such as, for example, the cancer, are readily measurable by routine procedures familiar to a physician.

[0037] “Therapeutically effective amount” is intended to refer to the level or amount of agent that is aimed at, without causing significant negative or adverse side effects to the target, (1) delaying or preventing the onset of a disease, such as, for example, a cancer; (2) slowing down or stopping the progression, aggravation, or deterioration of one or more symptoms of a disease, such as, for example, a cancer; (3) bringing about ameliorations of the symptoms of a disease, such as, for example, a cancer; (4) reducing the severity or incidence of a disease, such as, for example, a cancer; or (5) preventing disease formation, such as, for example, a cancer formation. The therapeutically effective amount may be administered prior to the onset of disease formation, such as, for example, a cancer formation, for a prophylactic or preventive action.

[0038] “Variant” refers to a polynucleotide that differs from a reference polynucleotide, but retains essential properties. A typical variant of a polynucleotide differs in nucleotide sequence from another, reference polynucleotide. Changes in the nucleotide sequence of the variant may or may not alter the amino acid sequence of a polypeptide encoded by the reference polynucleotide. Nucleotide changes may result in amino acid substitutions, additions, deletions, fusions and truncations in the polypeptide encoded by the reference sequence. A variant of a polynucleotide may be a naturally occurring such as an allelic variant. DETAILED DESCRIPTION

[0039] This invention discloses a method for predicting a risk of disease relapse in a subject previously treated for said disease, comprising the steps of:

(iii) based on the sequencing performed at step (ii), detecting the common cfDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and

(iv) concluding that the subject is at risk of disease relapse if at least one common cfDNA sequence is detected at step (iii).

[0040] In one embodiment, the disease is cancer, and the cfDNA molecule is a circulating tumor DNA (ctDNA) molecule. Thus, the present invention further discloses a method for predicting a risk of cancer relapse in a subject previously treated for cancer, comprising the steps of:

(i) isolating cell-free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said ctDNA molecules;

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i) and detecting the tumor circulating DNA (ctDNA); (iii) based on the sequencing performed at step (ii), detecting the common ctDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and

(iv) concluding that the subject is at risk of cancer relapse if at least one common ctDNA sequence is detected at step (iii).

[0041] The present invention further discloses a method for predicting a risk of cancer relapse in a subject previously treated for cancer, comprising the steps of:

(iii) isolating germline DNA molecules from a tissue sample, preferably from a whole blood sample and amplifying and sequencing said germline DNA molecules;

(iv) based on the sequencing performed at step (ii) and (iii), identifying ctDNA sequences among said cfDNA sequences;

(v) based on the identification of ctDNA sequences performed in step (iv), detecting the common ctDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and

(vi) concluding that the subject is at risk of cancer relapse if at least one common ctDNA sequence is detected at step (v). [0042] The present invention relates to a method for predicting a risk of cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(ii) sequencing the cfDNA molecules isolated, and optionally amplified, at step (i);

[0043] This invention further discloses a method for detecting a minimal residual disease (MRD) in a subject previously treated for said disease, comprising the steps of:

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i); (iii) based on the sequencing performed at step (ii), detecting the common cfDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and

(iv) concluding that the subject has a MRD if at least one common cfDNA sequence is detected at step (iii). wherein the MRD corresponds to the presence of at least one disease- specific cell in said subject.

[0044] This invention further discloses a method for detecting a minimal residual disease (MRD) in a subject previously treated for said disease, comprising the steps of:

(v) based on the identification of ctDNA sequences performed in step (iv), detecting the common ctDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and (vi) concluding that the subject has a MRD if at least one common ctDNA sequence is detected at step (v). wherein the MRD corresponds to the presence of at least one disease- specific cell in said subject.

[0045] In one embodiment, the disease is cancer, and the cfDNA molecule is a circulating tumor DNA (ctDNA) molecule. Thus, the present invention further discloses a method for detecting a minimal residual disease (MRD) in a subject previously treated for cancer, comprising the steps of:

(i) isolating cell-free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said cfDNA molecules;

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step and detecting the tumor circulating DNA (ctDNA) (i);

(iii) based on the sequencing performed at step (ii), detecting the common ctDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment; and

(iv) concluding that the subject has a MRD if at least one common ctDNA sequence is detected at step (iii). wherein the MRD corresponds to the presence of at least one cancer cell in said subject.

[0046] This invention further relates to a method for detecting a minimal residual disease (MRD) in a subject previously treated for a cancer, comprising the steps of:

(iv) concluding that said subject has a MRD if at least one ctDNA sequence is detected at step (iii). wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is selected from the group consisting of: blood, plasma and serum.

[0047] In some embodiments, the subject was previously treated for a disease selected from the group comprising or consisting of cancer.

[0048] The invention further relates to a method for treating a cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

[0049] In some embodiments, the subject was previously treated for cancer. Non- limitative examples of cancer types include carcinoma, lymphoma, blastoma, sarcoma, and leukemia, preferably carcinoma.

[0050] In some embodiments, the cancer is characterized in that it induces the presence of at least one circulating tumor DNA (ctDNA) molecule in the systemic circulation of the subject.

[0051] In some embodiments, the subject was previously treated for a cancer selected from the group comprising or consisting of carcinoma, lymphoma, blastoma, sarcoma, liposarcoma, neuroendocrine tumor, mesothelioma, schwannoma, meningioma, melanoma, leukemia, lymphoid malignancy, squamous cell cancer, epithelial squamous cell cancer, epithelial squamous cell cancer, lung cancer, small cell lung cancer, nonsmall cell lung cancer, adenocarcinoma of the lung, squamous carcinoma of the lung, cancer of the peritoneum, a hepatocellular cancer, a gastric or stomach cancer, a gastrointestinal cancer, pancreatic cancer, brain cancer, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, rectal cancer, colorectal cancer, endometrial or uterine carcinoma, salivary gland carcinoma, kidney and renal cancer, prostate cancer, vulval cancer, thyroid cancer, hepatic carcinoma, anal carcinoma, uterine cancer, penile carcinoma, testicular cancer, esophageal cancer, biliary tract cancer and head and neck cancer.

[0052] In some embodiments, the subject was previously treated for a head cancer or neck cancer. In some embodiments, the subject was previously treated for a head cancer. In some embodiments, the subject was previously treated for a neck cancer.

[0053] In some embodiments, the subject was previously treated for a locally advanced squamous cell carcinoma of the head cancer or neck cancer. In some embodiments, the subject was previously treated for a locally advanced squamous cell carcinoma of the head cancer. In some embodiments, the subject was previously treated for a locally advanced squamous cell carcinoma of the neck cancer.

[0054] In some embodiments, the subject was treated for locally advanced squamous cell carcinoma of the head cancer or neck cancer (SCCHN); in particular stage III or stage IV locally advanced SCCHN.

[0055] In some embodiments, the subject was previously treated for a virus induced cancer or virus-dependent cancer.

[0056] In some embodiments, the subject was previously treated for a virus induced cancer or virus dependent cancer, such as HPV16.

[0057] In some embodiments, the subject previously treated for a disease received at least one treatment suitable to treat said disease. In some embodiments, the subject is considered as treated or cured, the subject does not show clinical symptoms of the disease, and/or the subject is negative for routine clinical tests.

[0058] In some embodiments, the subject previously treated for cancer received at least one cancer treatment. In some embodiments, the subject previously treated for cancer received at least one cancer treatment selected from the group comprising or consisting of anticancer agents, anticancer peptides, anticancer nucleic acids and vectors, irradiation, immuno-oncotherapy, immune checkpoint inhibitors, cell therapy, and surgery. [0059] In some embodiments, the subject previously treated for cancer received at least one cancer treatment selected from the group comprising or consisting of anticancer agents, irradiation, and surgery.

[0060] In some embodiments, the subject previously treated for cancer may have received at least one anticancer agent. In some embodiments, anticancer agents are pharmaceutical, drugs and/or physiologically acceptable chemicals. In some embodiments, the subject previously treated for cancer may have received at least one therapeutically effective dose of said at least one anticancer agent, more preferably the subject previously treated for cancer received therapeutically effective doses of said at least one anticancer agent. In one embodiment, cancer is not detectable by conventional means in said subject.

[0061] Anticancer agents are known from the state of the art. Non-limitative examples of anticancer agents include acalabrutinib, alectinib, alemtuzumab, anastrozole, avapritinib, avelumab, belinostat, bevacizumab, bleomycin, blinatumomab, bosutinib, brigatinib, carboplatin, carmustine, cetuximab, chlorambucil, cisplatin copanlisib, cytarabine, daunorubicin, decitabine, dexamethasone, docetaxel, doxorubicin, encorafenib, erdafitinib, etoposide, everolimus, exemestane, fludarabine, 5-fluorouracil, gemcitabine, ifosfamide, imatinib Mesylate, leuprolide, lomustine, mechlorethamine, melphalan, methotrexate, mitomycin, nelarabine, paclitaxel, pamidronate, panobinostat, pralatrexate, prednisolone, ofatumumab, rituximab, temozolomide, topotecan, tositumomab, trastuzumab, vandetanib, vincristine, vorinostat, zanubrutinib, and the likes.

[0062] In some embodiments, the anticancer agents are Platine salt and/or 5-fluorouracil (5FU).

[0063] In some embodiments, the subject previously treated for cancer received at least one anticancer peptide. As used herein, an “anticancer peptide” refers and interchangeably to a peptide, polypeptide or protein, optionally comprising one or more post-translational modifications, that exerts an anticancer effect. In some embodiments, the anticancer peptide is administered along with at least one pharmaceutically acceptable excipient, solvent, or carrier. In a preferred embodiment, the subject previously treated for cancer received at least one therapeutically effective dose of said at least one anticancer peptide, more preferably the subject previously treated for cancer received therapeutically effective doses of said at least one anticancer peptide. In some embodiments, cancer is not detectable by conventional means in said subject.

[0064] Typically, an anticancer peptide comprises the sequence, or a fragment of the sequence, of an epitope of a cancer-associated protein or tumor-associated protein, preferably an immunogenic epitope. Alternatively, the anticancer peptide mimics an epitope of a cancer-associated protein or tumor-associated protein, preferably an immunogenic epitope. In some embodiments, the anticancer peptide is used as an anticancer and/or antitumor vaccine (z.e., preventive vaccine and/or therapeutic vaccine).

[0065] In some embodiments, the subject previously treated for cancer received at least one anticancer nucleic acid, or vector comprising or encoding thereof. Examples of anticancer nucleic acids comprise, but are not limited to, siRNA, sisiRNA, shRNA, asiRNA, aiRNA, miNRA, pre-miRNA, asDNA, and the like. Examples of anticancer nucleic acids comprise, but are not limited to, plasmids, fosmids, cosmids, artificial chromosomes, nanoparticles, liposomes, protein-nucleic acid complex, integrative viral vector, non-integrative viral vector, and the like. In some embodiments, the anticancer nucleic acid, or vector comprising or encoding thereof is administered along with at least one pharmaceutically acceptable excipient, solvent, or carrier. In a preferred embodiment, the subject previously treated for cancer received at least one therapeutically effective dose of said at least one anticancer nucleic acid, or vector comprising or encoding thereof, more preferably the subject previously treated for cancer received therapeutically effective doses of said at least one anticancer nucleic acid, or vector comprising or encoding thereof. In one embodiment, cancer is not detectable by conventional means in said subject.

[0066] In some embodiments, the subject previously treated for cancer received at least one immune checkpoint inhibitor. Non-limitative examples of immune checkpoint inhibitor include molecules targeting inducible T Cell costimulator (ICOS), programmed cell death protein- 1, ligand of PD-1, NKG2A, cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T cell immunoglobulin and mucin-domain containing-3 (TIM-3), CD112R, V-domain Ig suppressor of T cell activation (VISTA), lymphocyte- activation gene 3 (LAG-3), B- and T-lymphocyte attenuator (BTLA), T cell immunoglobulin and ITIM domain (TIGIT), B7 homolog 3 protein (B7-H3), Sialic acid binding Ig-like lectin 15 (Siglec-15), glucocorticoid-induced TNFR-related protein (GITR), cytokine-inducible SH2-containing protein (CISH), or combination thereof. In a preferred embodiment, the subject previously treated for cancer received at least one therapeutically effective dose of said at least one immune checkpoint inhibitor, more preferably the subject previously treated for cancer received therapeutically effective doses of said at least one immune checkpoint inhibitor. In one embodiment, cancer is not detectable by conventional means in said subject.

[0067] In some embodiments, the subject previously treated for cancer may have received at least one dose of irradiation (z.e., radiotherapy) to treat cancer. In some embodiments, the radiotherapy is an eternal beam radiation therapy or an internal radiation therapy. Typically, radiation doses used for radiotherapy are known from the art and may be adapted by a health practitioner depending on, e.g., tumor size/volume, and may range from 10 Gy to 100 Gy, although lower doses or higher doses may be used depending on the cancer to be treated. In a preferred embodiment, the subject previously treated for cancer received at least one therapeutically effective dose of said at least one radiotherapy, more preferably the subject previously treated for cancer received therapeutically effective doses of said at least one radiotherapy. In one embodiment, cancer is not detectable by conventional means in said subject.

[0068] In some embodiments, the radiotherapy is performed on tumor and/or on lymph nodes.

[0069] In some embodiments, the subject previously treated for cancer may have received at least one surgery. It will be apparent that a health practitioner may adapt the surgery depending on the type of cancer, in particular the type of tumor. The surgery may be, non-imitatively, curative surgery, preventive surgery, staging surgery, debulking surgery, palliative surgery, supportive surgery, restorative surgery, and/or combination thereof. [0070] In some embodiments, when the subject was previously treated for head and neck cancer, the treatment was anticancer agents, irradiations and/or surgery.

[0071] As used herein, “disease relapse” means that a patient that has undergone a curative-intent treatment starts to develop the disease again. The relapse may occur after any given amount of time after the end of the treatment.

[0072] As used herein, “cancer relapse” means that a patient that has undergone a curative-intent treatment start to develop cancer again, or in other words that cancer cells are proliferating again in their body. In some embodiments, patient with cancer relapse are patient did not complete their remission. In some embodiments, the term “relapse” refers also to “recidivism” or “recurrence” and can be used interchangeably throughout the present application.

[0073] In some embodiments, the method is used when a subject has undergone a curative treatment. In some embodiments, the method is used when a subject has undergone an anti-cancer treatment.

[0074] In some embodiments, the method is used when a subject is considered clinically treated for its disease, preferably cancer.

[0075] In some embodiments, the method is used when a subject is considered clinically treated for its disease, preferably cancer, meaning that cells that are specific for the disease (e.g., cancer cells, infected cells, microorganisms, and the like), preferably cancerous cells, are not detectable clinically nor with imagery.

[0076] In some embodiments, the method is used when a subject is considered clinically treated for the disease, preferably cancer, in order to predict the risk of relapse. In some embodiments, the method is used when a subject is considered clinically treated for its disease, preferably cancer, in order to detect MRD.

[0077] In some embodiments, the method is used when a subject is reminiscent. In some embodiments, the method is used when the disease, preferably cancer, is in remission. [0078] In some embodiments, the disease relapse, preferably cancer relapse occurs one week, one month, 6 months, a year, 2 years, or more, after the end of the treatment. In some embodiments, the disease, preferably cancer, remains “dormant” during one week, one month, 6 months, a year, 2 years, or more, after the end of the treatment.

[0079] In a preferred embodiment, the disease relapse occurs at least 2 years after the end of the treatment.

[0080] In one embodiment, disease relapse, preferably cancer relapse occurs at the same location (z.e., organ or tissue) than the one where it first appeared in the organism of the subject. In another embodiment, disease relapse, preferably cancer relapse occurs at a different location (z.e., organ or tissue) than the one where it first appeared in the organism of the subject.

[0081] In some embodiments, cancer relapse occurs when metastases (z.e., metastatic cells) are present in the organism of the subject.

[0082] In some embodiments, minimal residual disease (MRD) refers to disease- specific cells, preferably cancer cells remaining after treatment that may not be detected directly by classical analytical technics, such as scans or clinical evaluations. In some embodiments, MRD reflects the presence of a low number of disease- specific cells, preferably cancer cells in the body after cancer treatment, wherein “low number” means a number of cancer cells below the detection threshold and/or sensitivity threshold of classical analytical technics. In some embodiments, an MRD positive test result means that the disease was still detected after treatment. In some embodiments, an MRD negative result means that no disease was detected after treatment.

[0083] It will be apparent to the person skilled in the art that disease- specific cells, preferably cancer cells subsisting after treatment may proliferate again and cause a relapse. Therefore, in some embodiments, MRD causes cancer relapse. In some embodiments, MRD is associated with cancer relapse. In some embodiments, MRD is a risk factor for cancer relapse. [0084] It is to be understood that since classical analytical methods, such as scans or clinical evaluation may not detect MRD because the number of cells is too low to be detected by these techniques (z.e., below detection threshold), the use of these techniques may yield a false MRD negative result (z.e., the MRD is present but is not detected). The present invention takes advantage of the specific detection of circulating nucleic acid molecules, in particular DNA (z.e., cell free DNA, cfDNA), preferably tumor circulating DNA (ctDNA). Indeed, the detection of specific ctDNA molecules in the systemic circulation (z.e., blood) of the subject can be performed with a better sensitivity (i.e., with a lower detection threshold) compared to classical methods cited hereinabove. Hence, in some embodiments, the methods of the present invention enable a more robust and reliable detection of MRD. In some embodiments, the methods of the present invention enable a more robust and reliable detection of cancer relapse and/or risk of cancer relapse.

[0085] In some embodiments, the method of the invention does not comprise a step of collecting one or more biological sample from the subject.

[0086] In some embodiments, the method of the invention further comprises a step of obtaining at least one biological sample from the subject before treatment, and at least one biological sample from the subject after treatment

[0087] In some embodiments, the method of the invention further comprises a step of obtaining at least one biological sample, preferably at least one biological sample (whole blood) before, during, and/or after the treatment.

[0088] As used herein, “at least one” means 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more samples.

[0089] As used herein, “before treatment” means any time before the start of the anticancer treatment, e.g., 1 hour, 1 week, 1 month, or more, and after cancer diagnosis wherein “start of the anticancer treatment” refers to the first administration of the anticancer treatment.

[0090] As used herein, “after treatment” means any time after the end of the anticancer treatment, e.g., 1 week, 1 month, 1 year, or more, preferably at least one week, more preferably at least 12 weeks, even more preferably between 1 week and 12 weeks, wherein “end of the anticancer treatment” refers to the last administration of the anticancer treatment.

[0091] In some embodiments, the at least one biological sample collected after treatment is collected at least one week after the end of the treatment.

[0092] In some embodiments, the at least one biological sample collected after treatment is collected at least 2 weeks, 3 weeks, 4 weeks, 5 weeks, 6 weeks, 7 weeks, 8 weeks, 9 weeks, 10 weeks, 11 weeks, or more, after the end of the treatment.

[0093] In some embodiments, the at least one biological sample collected between 1 week and 12 weeks after the end of the treatment.

[0094] In some embodiments, the at least one biological sample collected after treatment is collected at least 12 weeks after the end of the treatment.

[0095] In some embodiments, the at least one biological sample collected after treatment is collected at least one week after the end of said treatment, preferably at least 12 weeks after the end of said treatment.

[0096] In some embodiments, the at least one biological sample collected after treatment is collected at least 24 weeks, 36 weeks, 48 weeks, or more, after the end of the treatment.

[0097] In some embodiments, more than one biological sample are collected before the treatment. As used herein, “more than one” means 2, 3, 4, 5, 6, 7, 8, 9, 10, or more.

[0098] In one embodiment, the samples are collected at a regular frequency. In certain embodiments, the samples are collected every day, every 2 days, every 3 days, every week, every 2 weeks, every 3 weeks, every month, every 2 months, every 3 months, every 4 months, every 5 months, every 6 months, or every year. In another embodiment, the samples are collected at a variable frequency, preferably according to health practitioner’ s decisions.

[0099] In some embodiments, more than one biological sample are collected after the treatment. As used herein, “more than one” means 2, 3, 4, 5, 6, 7, 8, 9, 10, or more. In some embodiments, more than one biological sample are collected after the treatment with the aim of performing a follow-up of MRD and/or the risk of cancer relapse. In some embodiments, more than one biological sample are collected after the treatment with the aim of detecting variations in MRD and/or the risk of cancer relapse. In one embodiment, the samples are collected at a regular frequency. In certain embodiments, the samples are collected every day, every 2 days, every 3 days, every week, every 2 weeks, every 3 weeks, every month, every 2 months, every 3 months, every 4 months, every 5 months, every 6 months, or every year. In another embodiment, the samples are collected at a variable frequency, preferably according to health practitioner’s decisions.

[0100] In some embodiments, the methods of the invention further comprise obtaining at least one biological sample from the subject during the treatment of the subject, such as, for example, during the anticancer treatment.

[0101] In some embodiments, the at least one biological sample may be obtained from any organ or tissue from the subject.

[0102] In some embodiments, the at least one biological sample is a body fluid. In some embodiments, the at least one biological sample is selected in the group comprising or consisting of cerebral spinal fluid, smear fluid, cyst fluid, pancreatic fluid, pleural fluid, intestinal fluid, urine, blood, plasma and serum, preferably blood, more preferably plasma.

[0103] In some embodiments, the at least one biological sample is selected in the group comprising or consisting of blood, plasma and serum.

[0104] In one embodiment, the at least one biological sample is blood.

[0105] In a preferred embodiment, the at least one biological sample is plasma. In some embodiments, the plasma sample is obtained from a blood sample of a subject after centrifugation.

[0106] In one embodiment, the at least one biological sample is serum. In some embodiments, the serum sample is obtained from a blood sample of a subject after centrifugation. [0107] In some embodiments, the at least one biological sample is obtained by conventional means, i.e., hypodermic needle and syringe, catheter, vacuum extraction systems, and the like.

[0108] In some embodiments, the at least one biological sample is obtained at a volume of 1 inL. 2 niL. 3 ml. 4 niL. 5 inL. 6 ml. 7 ml. 8 inL. 9 inL. 10 ml. 20 inL. 30 inL. 40 mL, 50 mL, 60 mL, 70 mL, 80 mL, 90 mL, 100 mL, or more, per biological sample, preferably between 4 mL and 6 mL per biological sample.

[0109] In some embodiments, the at least one biological sample is not from a solid tissue or organ. In some embodiments, the biological sample is not from a biopsy.

[0110] In one embodiment, the at least one biological sample before treatment and the biological sample after treatment are extracted from the same organ or tissue, preferably they are of the same type, (e.g., both are plasma samples).

[0111] In some embodiments, the at least one biological sample is treated to extract cfDNA molecules. Methods to prepare biological samples for cfDNA extraction are known in the art. Kits to perform such extraction are commercially available, such as Promega® Maxwell® RSC ccfDNA LV plasma kit, for example.

[0112] In some embodiments, the at least one biological sample are stored for at least one week, one month, one year, or more. Suitable methods to store biological samples are known in the art, typically the biological samples are stored at -80°C, in a freezer or in liquid nitrogen, or in liquid nitrogen vapors.

[0113] In some embodiments, the methods of the invention further comprise obtaining at least one biological sample from a any tissue or fluid sample from the subject, preferably a whole blood of the patient in order to extract germline or genomic DNA. In some embodiments, the germline or genomic DNA sample is collected before, during and/or after the treatment.

[0114] In some embodiments, step (iii) of the methods as described herein, further comprises comparing aligned sequence data obtained from germline DNA. [0115] In some embodiments, the at least one biological sample comprises biomarkers allowing the detection of cancer relapse. In some embodiments, the at least one biological sample comprises biomarkers allowing the detection of MRD.

[0116] In some embodiment, the biomarkers in the at least one biological sample are tumor- specific or tumor-derived molecules. In another embodiment, the biomarkers in the at least one biological sample are molecules specific and/or derived from a pathogen and/or an infected cell.

[0117] In some embodiments, the biomarkers in the at least one biological sample are tumor- specific molecules.

[0118] According to the present invention, the biomarker is a nucleic acid molecule.

[0119] As used herein, nucleic acid molecule refers to either desoxyribonucleic acid (for example, but not limited to, circulating DNA, complementary DNA or genomic DNA) or ribonucleic acid (for example, but not limited to, a mRNA, miRNAs, cRNA, IncRNA, or tRNA). The nucleic acid molecule can be single- stranded or double- stranded. In a preferred embodiment, the nucleic acid molecule is DNA.

[0120] In some embodiments, the biomarker is a genomic nucleic acid molecule.

[0121] In some embodiments, the biomarker is a DNA molecule, preferably a cell free DNA (cfDNA) molecule, even more preferably a circulating tumor DNA (ctDNA) molecule.

[0122] In some embodiments, the cfDNA or the ctDNA molecule is isolated from the biological sample, preferably from a plasma sample. In some embodiments, the plasma sample comprise at least one cfDNA or ctDNA molecule, preferably multiple copies of at least one cfDNA or ctDNA molecule.

[0123] In some embodiments, the cfDNA or ctDNA molecule comprises at least one mutation compared to sequences found in normal or healthy cell, preferably from the same subject. [0124] As used herein, the term “mutation” refers to a difference in a nucleotide sequence (e.g., DNA or RNA) in a tumor cell compared to a healthy cell from the same individual.

[0125] As used herein, the “genetic alteration” may be a single nucleotide polymorphism (SNP), INDEL, single nucleotide variants (mutations), substitutions, duplications, insertions, deletions, gene copy number variations, and structural variants, including inversions and translocations, gene fusions or another genetic alteration of interest.

[0126] In some embodiments, the cfDNA or ctDNA molecule mutation is selected from the group comprising or consisting of inversion, substitution, deletion, insertion, chromosomal rearrangement, frameshift mutation and the like.

[0127] In some embodiments, the cfDNA or ctDNA reflects the mutations associated to the tumor in a cancer patient.

[0128] In some embodiments, the cfDNA or ctDNA molecule comprises a sequence of at least one cancer-related gene, and comprises at least one mutation in said at least one cancer-related genes.

[0129] In some embodiments, the cfDNA or ctDNA molecule comprises a nucleic acid sequence of any one of the genes selected in the group comprising AKT1, ALK, APC, AR, ARAF, ARID 1 A, ARID2, ATM, B2M, BCL2, BCOR, BRAF, BRCA1, BRCA2, CARD11, CASP8, CBFB, CCND1, CDH1, CDK4, CDKN2A, CIC, CREBBP, CTCF, CTNNB1, DICER1, DIS3, DNMT3A, EGFR, EIF1AX, E6, E7, EP300, EPHA2, ERBB2, ERBB3, ERCC2, ESRI, EZH2, FAT1, FBXW7, FGFR1, FGFR2, FGFR3, FGFR4, FEG, FET3, FOSE2, FOXA1, FOXE2, FOXO1, FUBP1, GAT A3, GNA11, GNAQ, GNAS, H3F3A, HIST1H3B, HEA-A, HRAS, HRNR, IDH1, IDH2, IKZF1, INPPE1, JAK1, KDM6A, KEAP1, KIT, KMT2D, KNSTRN, KRAS, KTR5, MAP2K1, MAPK1, MAX, MED 12, MET, MLH1, MSH2, MSH3, MSH6, MTOR, MYC, MYCN, MYD88, MYODI, NECAB1, NF1, NFE2L2, NOTCH1, NSD1, NRAS, NTRK1, NTRK2, NTRK3, NUP93, PAK7, PDGFRA, PIK3CA, PIK3CB, PIK3R1, PIK3R2, PMS2, POLE, PPP2R1A, PPP6C, PRKCI, PSIP1, PTCHI, PTEN, PTPN11, RAC1, RAFI, RBI, RET, RHOA, RIT1, ROS1, RRAS2, RXRA, SETD2, SF3B1, SMAD3, SMAD4, SMARCA4, SMARCB1, S0S1, SPOP, STAT3, STK11, STK19, TCF7L2, TGFBR1, TGFBR2, TP53, TP63, TSC1, TSC2, U2AF1, VHL, XP01, TERT, or variant and/or a fragment thereof.

[0130] In one embodiment, the cfDNA or ctDNA molecule comprises a nucleic acid sequence of at least one gene selected in the group comprising or consisting of AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof, wherein E6 and E7 are two HPV16 oncogenes.

[0131] In one embodiment, the cfDNA or ctDNA molecule comprises a nucleic acid sequence of at least one gene selected in the group comprising or consisting of AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof, wherein E6 and E7 are two HPV16 oncogenes, wherein the gene panel is used for the detection of head and neck cancer, preferably SSCHN.

[0132] In some embodiment, the cfDNA or ctDNA molecule is used to identify head and neck cancer, preferably SCCHN and comprises a nucleic acid sequence of any one of the genes selected in the group comprising AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof.

[0133] In some embodiments, the ctDNA sequence comprises a nucleic acid sequence of any one of the genes selected in the group consisting of: AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof.

[0134] In some embodiments, the ctDNA sequence comprises a nucleic acid sequence selected from E6 or E7. [0135] In some embodiments, the cfDNA or ctDNA molecule or sequence comprises at least one mutation in at least one cancer-related genes selected in the list disclosed herein above, or variant and/or a fragment thereof.

[0136] In some embodiments, the cfDNA or ctDNA molecule comprises at least one mutation in at least one cancer-related genes selected in the list disclosed herein above, or variant and/or a fragment thereof.

[0137] In one embodiment, the method of the present invention comprises a step of amplifying cfDNA or ctDNA molecules, such as, for example, using Polymerase Chain Reactions (PCR). In some embodiments, the cfDNA or ctDNA molecules are amplified using DNA primers specifically designed to amplify at least one cancer-related gene selected in the list disclosed herein above.

[0138] In some embodiments, cfDNA molecules primers and ctDNA molecule primers are locus- specific primers chosen so as to identify a specific mutation or variant of a nucleotide sequence (i.e., a genetic marker) that may be present in the biological sample, wherein said mutation is indicative of the disease for which the patient has been treated.

[0139] According to the present invention, the method of the present invention comprises a step of sequencing the cfDNA molecules isolated from the samples obtained from the subject, and a step of identifying ctDNA sequences that is identical in the sample obtained from the subject before treatment and in the sample obtained from the subject after treatment i.e., detecting common ctDNA sequences). In some embodiments, two ctDNA sequences are considered identical if they share the same nucleic acid sequence.

[0140] In some embodiments, a ctDNA molecule from at least one biological sample collected before treatment and a ctDNA molecule from at least one biological sample collected after treatment are considered identical if they share the same nucleic acid sequence.

[0141] In some embodiments, the cfDNA molecules isolated from the samples are subjected to high-throughput sequencing, commonly referred to as DNA-seq. This method is also called next-generation sequencing (NGS). [0142] NGS is a method known to the skilled in the art. Briefly, this method comprises steps of library preparation, DNA amplification by polymerase chain reaction, followed by a step of sequencing using primers thereby producing reads. In some embodiments, reads are next aligned to a pre-sequenced reference genome (such as, for example, the human genome reference). In some embodiments, the number of reads mapped to a gene quantifies the expression level. In some embodiments, quality and specificity assessment may be performed throughout the experiment.

[0143] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of genes of interest.

[0144] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of cancer-related genes. In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of head and neck cancer related genes.

[0145] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of SCCHN- related genes. Examples of SCCHN related gene include, but are not limited to AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7.

[0146] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of virus-induced cancer dependent related genes. In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of HPV16-related genes.

[0147] In some embodiments, step (iii) of the methods as described herein comprises the following steps: - performing sequence alignment on genome, preferably on human genome and optionally on virus-induced cancer genome;

- measuring a cfDNA variant allele frequency (VAF) in the biological samples.

[0148] According to the invention, a cfDNA variant is selected when: the cfDNA from said at least one biological sample collected before treatment is characterized in that: (a) at least three reads are detected for said cfDNA; (b) the reads are aligned in forward and reverse; and (c) the cfDNA is present at less than 1% of healthy individuals in public sequence databases; and/or the cfDNA from said at least one biological sample collected after treatment is characterized in that: (a) said cfDNA is also identified in the at least one biological sample collected before treatment; (b) at least three reads are detected for said cfDNA.

[0149] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of cancer-related genes selected in the list comprising or consisting of A JUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof. In one embodiment, these genes are particularly interesting for the identification of head and neck cancer such as SSCHN.

[0150] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the exons of the following 26 genes: AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7. In one embodiment, these genes are particularly interesting for the identification of head and neck cancer such as SSCHN.

[0151] In some embodiments, SCCHN is HPV-positive SCCHN, preferably HPV-16 positive SCCHN.

[0152] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing of the genes of E6 and E7.

[0153] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising a first step of analyzing virus-induced cancer gene, following by a second step of analyzing cancer-related genes.

[0154] In some embodiments, the method of the present invention comprises a targeted NGS step, comprising DNA amplification and sequencing, in a first step, analyzing the 2 HPV16 genes E6, E7, and in a second step, analyzing at least one cancer-related gene (preferably all the cancer related genes) selected in the list comprising or consisting of AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, or variant and/or a fragment thereof.

[0155] In some embodiments, the sequencing generates Unique Molecular Identifiers (UMIs) on the sequences. [0156] In some embodiments, the sequencing data recovered at step (ii) of the method of the invention are reverted as “raw data”. Throughout the text, “raw data” may also be referred as “raw files” or “raw data files” and can be used interchangeably.

[0157] In some embodiments, “raw data files” correspond to a file format selected from a group comprising or consisting of “.fastq files” or “.fasta files”.

[0158] In certain embodiments, the sequencing data recovered at step (ii) of the method of the invention are “.fastq files”.

[0159] In some embodiment, the sequencing data recovered at step (ii) of the method of the invention correspond to the reads after sequencing.

[0160] In some embodiments, the method of the present invention comprises a step of bioinformatically analyzing the sequencing data recovered at step (ii). In one embodiment, the bioinformatic analysis is performed using a bioinformatic pipeline. In one embodiment, the step of bioinformatically analyzing the sequencing data obtained at step (ii) is referred as step (iii) according to the method of the invention.

[0161] It will be understood that in step (iii), the detection and identification of cfDNA sequences may comprise one or more steps of bioinformatical analysis that form a bioinformatic pipeline, as defined hereinbelow.

[0162] In some embodiments, the bioinformatic pipeline according to step (iii) of the method of the invention uses public bioinformatic tools such as software, public platform, package, algorithm, and the like. Therefore, it is assumed the information or particularities related to these bioinformatic tools are fully available and detailed on internet to enable the skilled artisan of the field to performed the invention.

[0163] In some embodiments, the bioinformatic analyses according to step (iii) of the method of the invention comprises analyses of the sequencing data recovered at step (ii) with the public platforms Kraken2 and/or Highlander.

[0164] It is well known in the field that Kraken2 is a taxonomic sequence classifier that assigns taxonomic labels to DNA sequences. Kraken examines the K-mers within a query sequence and uses the information within those K-mers to query a database. That database maps K-mers to the lowest common ancestor (LCA) of all genomes known to contain a given K-mer. Other taxonomic assignation tools comprise but are not restricted to MetaPhlAn, Pavian, Krona.

[0165] It is well known in the field that, Highlander is a bioinformatics framework, developed by the Genomics and Bioinformatics platform, owned by the UCLouvain, and allows variant-annotation, filtering and visualization of Genomic data.

[0166] In some embodiments, the bioinformatic analyses according to step (iii) of the method of the invention comprises analyses of the sequencing data recovered at step (ii) in a first step with Kraken2 and a second step with Highlander.

[0167] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises analyses of the sequencing data recovered at step (ii) with Highlander.

[0168] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises analyses of the sequencing data recovered at step (ii) that does not use Kraken2.

[0169] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2. In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2 in biological sample collected before treatment.

[0170] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2 in order to map reads on the genome.

[0171] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2, in order to detect the presence of reads mapping virus-induced cancer genome. [0172] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2, in order to detect the presence of reads mapping the HPV16 genome.

[0173] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in using Kraken2 in order to detect the presence of reads whose sequences comprise or consist in E6 and E7 sequences.

[0174] In some embodiments, the detection of at least one read mapping the virus- induced genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample comprise ctDNA molecule related to virus dependent-cancer.

[0175] In some embodiments, the detection of at least one read mapping the HPV16 genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample comprise ctDNA molecule related to HPV16 dependent-cancer.

[0176] In some embodiments, the detection of at least one read mapping the virus- induced genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample is “virus-induced cancer positive”.

[0177] In some embodiments, the detection of at least one read mapping the HPV16 genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample is “HPV16 positive”.

[0178] In some embodiments, the presence of reads mapping virus-induced cancer genome means that the biological sample is “ctDNA positive”.

[0179] In some embodiments, the presence of reads mapping HPV-induced cancer genome, preferably HPV16 genome, means that the biological sample is “ctDNA positive”.

[0180] In some embodiments, the absence of read mapping the virus-induced genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample does not comprise ctDNA molecule related to virus-dependent-cancer (or virus induced cancer). [0181] In some embodiments, the absence of read mapping the HPV16 genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample does not comprise ctDNA molecule related to HPV16 dependent-cancer (or virus induced cancer).

[0182] In some embodiments, the absence of read mapping the virus-induced genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample is “virus-induced cancer negative”.

[0183] In some embodiments, the absence of read mapping the HPV16 genome by Kraken2 according to step (iii) of the method of the invention means that the biological sample is “HPV16 negative”.

[0184] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in alignment of the sequencing data recovered at step (ii) using alignment bioinformatic tools.

[0185] In some embodiments, the “aligned sequences files” or “aligned reads files” recovered after sequence alignment step according to step (iii) of the method of the invention are reverted as “.bam files”.

[0186] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in alignment of the sequencing data recovered at step (ii) to a large reference genome.

[0187] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises or consists in alignment of the sequencing data recovered at step (ii) to a human genome (Hg38).

[0188] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention is processed by the public platform Highlander.

[0189] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention is performed using a bioinformatic tools well known in the art. Non limitative examples include BBMap, BFAST, BLASTN, Bowtie, BWA (Burrows- Wheeler Aligner), CUSHAW, GEM, Gensearch NGS, NextGen Map, RMAP, and the like.

[0190] In one embodiment, the step of sequence alignment of the bioinformatical analyses according to step (iii) of the method of the invention is performed using BWA (Burrows-Wheeler Aligner).

[0191] In some embodiments, the aligned sequences files recovered according to step (iii) of the method of the invention are further processed by bioinformatic tools well known in the art.

[0192] In some embodiments, the aligned sequences files recovered according to step (iii) of the method of the invention are then processed using fgbio tools and Picard in order to remove Unique Molecular Identifiers (UMIs). It will be apparent to the person skilled in the art that UMIs are a type of molecular barcoding that provides error correction and increased accuracy during sequencing. In some embodiments, UMIs reduce the rate of false-positive variant calls and increase sensitivity of variant detection.

[0193] In some embodiments, the aligned sequences files recovered according to step (iii) of the method of the invention are further analyzed or processed using bioinformatical tools in order to identify variant, such as single-nucleotide variants and small indels. In some embodiment, the step of identified variant in aligned sequences files is also called “variant calls”.

[0194] In some embodiments, the variant identification according to step (iii) of the method of the invention is processed in the germline DNA and/or the cfDNA aligned sequences files.

[0195] In some embodiments, the variant calls files (.vcf) recovered after the variant calls step according to step (iii) of the method of the invention are reverted as “.bam files”.

[0196] In some embodiments, the germline DNA aligned sequences files are processed to identify germline single-nucleotide variants and small indels. [0197] In some embodiments, the aligned sequences files recovered according to step (iii) of the method of the invention are processed using GATK 4.2 "BQSR for base quality score recalibration.

[0198] In some embodiments, GATK 4.2 "Haplotype Caller" is a variant caller, that identifies germline single-nucleotide variants and small indels, developed by the Broad Institute. In some embodiments, GATK 4.2 HaplotypeCaller is capable of calling SNPs and indels simultaneously via local de-novo assembly of haplotypes in an active region.

[0199] In some embodiments, the ctDNA aligned sequences files recovered according to step (iii) of the method of the invention are processed to identify somatic single- nucleotide variants (SNV) and small indels.

[0200] In some embodiments, the aligned sequences files recovered according to step (iii) of the method of the invention are processed using Mutect2.

[0201] In some embodiments, Mutec2 identifies somatic single-nucleotide variants (SNV) and small indels, developed by the Broad Institute. In some embodiments, Mutec2 is a caller that uses a Bayesian somatic genotyping and uses the assembly-based machinery of HaplotypeCaller. In some embodiments, Mutect2 has shown similar performance in identifying somatic SNVs and indels as other caller, such as but not only Stelka2, VarScan2.

[0202] In some embodiments, the variant calls files (.vcf) recovered from the variant identification analyses according to step (iii) of the method from the invention are annotated, imported and further analyzed using Highlander

[0203] In some embodiments, the cfDNA variant or ctDNA variants are SNVs.

[0204] In some embodiments, the SNVs recovered from the SNVs analyses according to step (iii) of the method of the invention are selected based on quality and selective filters filtering tools performed by Highlander.

[0205] In some embodiments, the quality and selective filters performed by Highlander comprise that the selected-SNVs meet one or more the following criteria: - should not be present in more than 1% across all healthy populations, according to the proportions reported in public databases such as Exac and Gnomad;

- should be present in a least one read in both orientations;

- should be on reads passing the quality filters from the GATK institution and Mutect2 pipeline (list of quality filters can be found on the internet);

- should not be present in an intronic section of the genome;

- should be present on reads not exceeding 100 base pairs;

- should be present in at least 3 reads; and/or

- should be present at a position where at least 500 reads are aligned.

[0206] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected-SNVs should not be present in more than 1% across all healthy populations, according to the proportions reported in public databases such as Exac and Gnomad.

[0207] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected-SNVs should be present in a least one read in both orientations.

[0208] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected SNVs should be on reads passing the quality filters from the GATK institution and Mutect2 pipeline (list of quality filters can be found on the internet).

[0209] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected SNVs should not be present in an intronic section of the genome. [0210] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected SNVs should be present on reads not exceeding 100 base pairs.

[0211] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected-SNVs should be present in at least 3 reads

[0212] In some embodiments, the quality and selective filters performed by Highlander consists in that the selected SNVs should be present at a position where at least 500 reads are aligned.

[0213] In some embodiments, the variant identification according to step (iii) of the method of the invention that passed the filtration steps are ctDNA.

[0214] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention comprises measuring the variant allele frequency (VAF) in the samples, as the number of reads harboring the selected SNV on the total number of reads at the same genomic position.

[0215] In some embodiments, the variant identification in germline DNA analyses are used as reference for the variant of ctDNA analyses.

[0216] In some embodiments, the method of the invention comprises comparing the results with public sequence databases, such as Exac or Gnomad.

[0217] In some embodiments, during the step of detecting the common cfDNA or ctDNA molecules between said at least one biological sample collected before treatment and said at least one biological sample collected after treatment, a quality control is performed for each cfDNA or ctDNA molecule, wherein: a cfDNA or ctDNA from the at least one biological sample collected before treatment is validated in the quality control if (a) at least 3 reads are detected for the ctDNA; (b) the reads are aligned in forward and reverse; (c) the reads passed quality filters; and (d) the - cfDNA or ctDNA variant is present at less than 1% of healthy individuals in public sequence databases; a cfDNA or ctDNA from the at least one biological sample collected after treatment is validated in the quality control if (a) the cfDNA or ctDNA was also identified in the at least one biological sample collected before treatment; (b) the cfDNA or ctDNA is present in the raw sequencing files (.bam); and (c) at least 3 reads are detected for the cfDNA or ctDNA.

[0218] In some embodiment, the quality control of the step of detecting the common cfDNA or ctDNA molecules between said at least one biological sample collected before treatment and said at least one biological sample collected after treatment is performed for each cfDNA or ctDNA molecule, wherein a cfDNA or ctDNA from the at least one biological sample collected before treatment is validated in the quality control if (a) at least 3 reads are detected for the cfDNA or ctDNA; (b) the reads are aligned in forward and reverse; (c) the reads passed quality filters; and (d) the cfDNA or ctDNA is present at less than 1% of healthy individuals in public sequence databases.

[0219] In some embodiment, the quality control of the step of detecting the common cfDNA or ctDNA molecules between said at least one biological sample collected before treatment and said at least one biological sample collected after treatment is performed for each cfDNA or ctDNA molecule, wherein a cfDNA or ctDNA from the at least one biological sample collected after treatment is validated in the quality control if (a) the cfDNA or ctDNA was also identified in the at least one biological sample collected before treatment; (b) the cfDNA or ctDNA is present in the raw sequencing files; and (c) at least 3 reads are detected for the cfDNA or ctDNA.

[0220] In some embodiments, the bioinformatic analyses according to step (iii) of the method of the invention comprise analyses using other available public bioinformatic tools such as software, public platform, package, algorithm, and the like.

[0221] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention is designed to discriminate wild-type sequences from mutated sequences at specific positions, preferably in cancer-related genes).

[0222] In some embodiments, the bioinformatical analyses according to step (iii) of the method of the invention further comprise detecting the common ctDNA variant isolated from said at least one biological sample collected before treatment and said at least one biological sample collected after treatment

[0223] In some embodiments, the step of detecting the common cfDNA sequences between said at least one biological sample collected before treatment and said at least one biological sample collected after treatment further comprises detecting variants of the cfDNA sequences.

[0224] In some embodiments, the step of detecting the common ctDNA sequences between said at least one biological sample collected before treatment and said at least one biological sample collected after treatment further comprises detecting variants of the ctDNA sequences.

[0225] In some embodiments, the method comprises detecting at least one variant of the cfDNA sequences in the at least one sample collected before treatment. In some embodiments, the method comprises detecting at least one variant of the cfDNA sequences in the at least one sample collect posttreatment.

[0226] In some embodiments, the method comprises detecting at least one variant of the ctDNA sequences in the at least one sample collected before treatment. In some embodiments, the method comprises detecting at least one variant of the ctDNA sequences in the at least one sample collect posttreatment.

[0227] In some embodiments, the method comprises comparing the at least one variant of the cfDNA detected in the at least one sample collected before and after treatment.

[0228] In some embodiments, the method comprises comparing the at least one variant of the ctDNA detected in the at least one sample collected before and after treatment.

[0229] In some embodiments, the patient is considered positive for MRD and/or at risk of disease relapse if at least one variant of at least one common ctDNA molecule is detected in the at least one sample collected after treatment compared to the at least one sample collected before treatment. [0230] In some embodiments, the patient is considered negative if no variant of at least one ctDNA sequence is detected in the post treatment sample compared to the pretreatment sample.

[0231] In some embodiments, the biological sample obtained after the treatment is considered “MRD-positive” if at least 1 selected SNV from the biological sample obtained before the treatment is found in the biological sample after the treatment.

[0232] In some embodiments, the biological sample obtained after the treatment is considered “MRD-negative” if at least 1 selected SNV from the biological sample obtained before the treatment is absent in the biological sample after the treatment.

[0233] In some embodiments, “MRD-positive” means the patient is at risk of disease relapse.

[0234] In some embodiments, “MRD-positive” means the patient is not at risk of disease relapse.

[0235] In some embodiments, statistical analyses are performed after bioinformatic analyses. The method for statistical analysis is well known in the field and routinely used by the skilled artisan.

[0236] In some embodiments, correlation analyses are performed after bioinformatic analyses. The method for correlation analyses is well known in the field and routinely used by the skilled artisan.

[0237] In one embodiment, the specific bioinformatic is split in two phases, wherein the first one is performed on the DNA extracted from the pre-treatment samples and the second one is performed on the DNA from the post-treatment samples. In one embodiment, it comprises the following steps:

1) On pre-treatment samples:

Raw data files from the sequencing of cfDNA molecules extracted from the pretreatment samples are processed using Kraken2, a public taxonomic assignation bio-informatic tool, to determine the presence of HPV16. If DNA molecules from the raw data files include reads whose sequences include E6 and E7, they will be mapped to the HPV16 genome by the taxonomic assignation tool Kraken2. If at least on read from the raw data files is mapped to the HPV16 genome, the pretreatment sample from with the raw data file contains this read will be considered “HPV16-positive”. If no read from a raw data file is mapped to the HPV16 genome using Kraken2 taxinomic assignation tool, the pre-treatment sample from which the raw data file corresponds will be considered “HPV 16-negative”. Raw data files from the “HPV 16-negative” pre-treatment samples, according to the analysis described in step (i), are aligned to human genome (Hg38) using alignment bioinformatic tools such as BWA (Burrows-Wheeler Aligner). Aligned reads are then submitted to GATK Haplotye Caller bio-informatic pipeline to single-out single-nucleotide variations (SNVs) comparing reads from cfDNA from pre-treatment samples and human genome reference.

In parallel, raw data files from cfDNA extracted from pre-treatment samples are aligned to the corresponding aligned raw data files from the gDNA extracted from a whole blood sample, using public bio-informatic pipelines such as MUTECT2. ctDNA SNVs are thus singled-out by comparing reads from the cfDNA extracted from the pre-treatment sample and the germline or genomic DNA extracted from the whole blood. Steps (ii) and (iii) are performed using an interactive bioinformatic framework, called Highlander, developed by the Genomics and Bioinformatics platform, owned by the UCLouvain.

(i) Using the Highlander filtering tools, tumor-SNVs are selected based on quality and selective filters comprising of: a. Selected-SNVs sould no be present in more than 1% across all healthy populations, according to the proportions reported in public databases such as Exac and Gnomad. b. Selected SNVs should be present in a least one read in both orientations c. Selected SNVs should be on reads passing the quality filters from the GATK institution and Mutect2 pipeline (list of quality filters can be found on the internet). d. Selected SNVs should not be present in an intronic section of the genome. e. Selected SNVs should be present on reads not exceeding lOObase pairs. f. Selected-SNVs should be present in at least 3 reads g. Selected SNVs should be present at a position where at least 500 reads are aligned.

(ii) All selected SNV s are checked manually to eliminate alignment errors . All selected SNVs are then considered tumor variants. The presence of tumor variants in the pre-treatment plasma sample is considered as presence of ctDNA in the pre-treatment plasma sample.

If samples are “HPV 16-positive” or “HPV-negative” with ctDNA detected, the pre-treatment plasma samples are considered “ctDNA positive”. ) On the cfDNA extracted from the post-treatment samples:

(i) If pre-treatment samples are considered “HPV16-positives”, raw data files from the corresponding sequencing of cfDNA molecules extracted from the post-treatment samples are processed using Kraken2, a public taxonomic assignation bio-informatic tool, to determine the presence of HPV16. If DNA molecules from the raw data files include reads whose sequences include E6 and E7, they are mapped to the HPV 16 genome by the taxonomic assignation tool Kraken2. If at least one read from the raw data files is mapped to the HPV 16 genome, the post-treatment sample from which the raw data file contains this read is considered “MRD-positive”. If no read from a raw data file is mapped to the HPV 16 genome using Kraken2 taxinomic assignation tool, the post-treatment sample from which the raw data file corresponds is considered “MRD-negative”.

(ii) If pre-treatment samples are “HPV 16-negative” and “ctDNA positive”, the presence of the same selected SNVs in the post- treatment samples is assessed following these steps: a. Position in the genome of the selected pre-treatment SNVs is determined. b. If the same selected SNVs are presented in the corresponding raw data files (.bam) from the cfDNA extracted from the posttreatment samples in at least 3 reads, the selected SNVs are considered as present in the post-treatment samples. The posttreatment plasma sample is considered “MRD-positive” if at least 1 selected SNV from the pre-treatment sample is found in the post-treatment sample. If selected SNVs are not present in the post-treatment sample, the post-treatment sample is considered “MRD-negative”.

[0238] This invention also discloses a method for treating a disease relapse in a subject previously treated for said disease and identified as at risk of disease relapse or in which a MRD has been detected, comprising the steps of:

(i) detecting MRD or predicting a risk of relapse of said disease, using a method as described herein;

(ii) administering to the subject a treatment for treating and/or preventing said disease if a risk of disease relapse or a MRD has been detected in said subject.

[0239] In one embodiment, the treatment step comprises administering to said subject at least one treatment for the disease. In one embodiment, the treatment step comprises administering to said subject at least one cancer treatment selected from the group comprising or consisting of anticancer agents, anticancer peptides, anticancer nucleic acids and vectors, irradiation, immuno-oncotherapy, immune checkpoint inhibitors, cell therapy, and surgery. In some embodiments, the at least one cancer treatment is supplemented with at least one adjuvant. Thus, in certain embodiments, the treatment step comprises administering to said subject a therapeutically effective dose of an anticancer agent.

[0240] Examples of cancer treatments are listed hereinabove. [0241] In one embodiment, the treatment step is the same treatment as the treatment previously received by the subject. In one embodiment, the cancer treatment is the same cancer treatment as the treatment previously received by the subject.

[0242] In one embodiment, the treatment step is different compared to the treatment previously received by the subject. In some embodiments, the duration, frequency and/or dosage of the treatment is different.

[0243] The present invention further discloses a bioinformatic pipeline adapted to implement the method of the present invention.

[0244] The present invention further relates to a computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of the methods as described herein.

[0245] The present invention further relates to a computer-readable medium having stored thereon the computer program as described herein.

[0246] The present invention further relates to a kit comprising means for implementing the methods of the invention.

[0247] The present invention relates to a method for predicting a risk of cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(iv) concluding that said subject is at risk of cancer relapse if at least one ctDNA sequence is detected at step (iii); wherein the cancer is squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is plasma.

[0248] The present invention further relates to a method for detecting a minimal residual disease (MRD) in a subject previously treated for a cancer, comprising the steps of:

(iv) concluding that said subject has a MRD if at least one ctDNA sequence is detected at step (iii). wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is plasma. [0249] The present invention further relates to a method for treating a cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(v) administering to said subject a treatment for treating and/or preventing said disease if said subject is detected as being at risk of said disease relapse at step (iv) wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is plasma.

[0250] This invention further relates to a computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of a method for predicting a risk of cancer relapse, in a subject previously treated for said cancer, comprising the steps of: (i) isolating cell free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said cfDNA molecules;

[0251] This invention further relates to a computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of a method for detecting a minimal residual disease (MRD) in a subject previously treated for a cancer, comprising the steps of:

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i); (iii) based on the sequencing performed at step (ii), detecting, among the cfDNA sequences, the circulating tumor DNA (ctDNA) sequences isolated from said at least one biological sample collected before treatment and the ctDNA sequences isolated from the at least one biological sample collected after treatment; and

(iv) concluding that said subject has a MRD if at least one ctDNA sequence is detected at step (iii). wherein said MRD corresponds to the presence of at least one cancer cell in said subject, the cancer being squamous cell carcinoma of the head cancer or neck cancer (SCCHN); and wherein the biological sample is plasma.

[0252] This invention further relates to a computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of a method for treating a cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

(iii) based on the sequencing performed at step (ii), detecting, among the cfDNA sequences, the circulating tumor DNA (ctDNA) sequences isolated from said at least one biological sample collected before treatment and the ctDNA sequences isolated from the at least one biological sample collected after treatment; (iv) concluding that said subject is at risk of cancer relapse if at least one ctDNA sequence is detected at step (iii); and

[0253] A computer-readable medium having stored thereon the computer program as described herein.

[0254] This invention relates to a method for predicting a risk of disease relapse, preferably cancer relapse, in a subject previously treated for said disease, comprising the steps of:

(ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i) and optionally detecting the tumor circulating DNA (ctDNA);

(iii) based on the sequencing performed at step (ii), detecting the common cfDNA sequences isolated from said at least one biological sample collected before treatment and said at least one biological sample collected after treatment; and (iv) concluding that said subject is at risk of disease relapse if at least one common cfDNA sequence is detected at step (iii).

[0255] This invention relates to a method for detecting a minimal residual disease (MRD) in a subject previously treated for a disease, preferably cancer, comprising the steps of:

(iii) based on the sequencing performed at step (ii), detecting the common cfDNA sequences isolated from said at least one biological sample collected before treatment and said at least one biological sample collected after treatment; and

(iv) concluding that said subject has a MRD if at least one common cfDNA sequence is detected at step (iii). wherein said MRD corresponds to the presence of at least one cancer cell in said subject.

[0256] This invention relates to a method for treating a disease relapse, preferably cancer relapse, in a subject previously treated for the disease, comprising the steps of:

(i) isolating cell free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying the cfDNA molecules; (ii) sequencing the cfDNA molecules isolated and optionally amplified at step (i) and optionally detecting the tumor circulating DNA (ctDNA);

(iii) based on the sequencing performed at step (ii), detecting the common cfDNA sequences isolated from the at least one biological sample collected before treatment and the at least one biological sample collected after treatment;

(iv) concluding that the subject is at risk of disease relapse if at least one common cfDNA sequence is detected at step (iii); and

(v) administering to the subject a treatment for treating and/or preventing said disease if the subject is detected as being at risk of the disease relapse at step (iv).

[0257] In some embodiments, the step (iii) further comprises comparing the aligned sequence data obtained from germline DNA sample obtained from whole blood sample.

[0258] In some embodiments, the step (iii) comprises one more of: performing sequence alignment on genome, preferably on human genome and optionally on virus-induced cancer genome identifying variants of the cfDNA, such as single-nucleotide variants (SNV) and small indels variants. selecting the cfDNA variants using selective and quality filters wherein the selected-cfDNA variants meet one or more the following criteria: should not be present in more than 1 % across all healthy populations, according to the proportions reported in public databases such as Exac and Gnomad should be present in a least one read in both orientations

- should be on reads passing the quality filters from the GATK institution and Mutect2 pipeline (list of quality filters can be found on the internet) should not be present in an intronic section of the genome should be present on reads not exceeding 100 base pairs should be present in at least 3 reads should be present at a position where at least 500 reads are aligned, measuring the cfDNA variant allele frequency (VAF) in the samples.

[0259] In some embodiments, the step (iii), a quality control is performed for each cfDNA molecule, wherein: a cfDNA from the at least one biological sample collected before treatment is validated in the quality control if (a) at least 3 reads are detected for the cfDNA; (b) the reads are aligned in forward and reverse; (c) the reads passed quality filters; and (d) the cfDNA is present at less than 1% of healthy individuals in public sequence databases; a cfDNA from the at least one biological sample collected after treatment is validated in the quality control if (a) the cfDNA was also identified in the at least one biological sample collected before treatment; (b) the cfDNA is present in the raw sequencing files; and (c) at least 3 reads are detected for the cfDNA.

[0260] In some embodiments, two cfDNA molecules are considered common if they share the same nucleic acid sequence.

[0261] In some embodiments, the at least one biological sample is selected from the group comprising blood, plasma and serum, preferably plasma.

[0262] In some embodiments, the at least one biological sample collected after treatment is collected at least one week after the end of the treatment, preferably at least 12 weeks after the end of the treatment.

[0263] In some embodiments, more than one biological sample are collected after the treatment.

[0264] In some embodiments, the subject was treated for a cancer selected from the group comprising carcinoma, lymphoma, blastoma, sarcoma, liposarcoma, neuroendocrine tumor, mesothelioma, schwannoma, meningioma, melanoma, leukemia, lymphoid malignancy, squamous cell cancer, epithelial squamous cell cancer, epithelial squamous cell cancer, lung cancer, small cell lung cancer, non- small cell lung cancer, adenocarcinoma of the lung, squamous carcinoma of the lung, cancer of the peritoneum, a hepatocellular cancer, a gastric or stomach cancer, a gastrointestinal cancer, pancreatic cancer, brain cancer, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, rectal cancer, colorectal cancer, endometrial or uterine carcinoma, salivary gland carcinoma, kidney and renal cancer, prostate cancer, vulval cancer, thyroid cancer, hepatic carcinoma, anal carcinoma, uterine cancer, penile carcinoma, testicular cancer, esophageal cancer, biliary tract cancer and head and neck cancer.

[0265] In some embodiments, the subject was treated for head cancer or neck cancer.

[0266] In some embodiments, the subject was treated for locally advanced squamous cell carcinoma of the head cancer or neck cancer (SCCHN).

[0267] In some embodiments, the cfDNA molecule used to identify SCCHN comprises a nucleic acid sequence of any one of the genes selected in the group comprising AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof.

[0268] The method also relates to a bioinformatic pipeline adapted for implementing the method according to the method of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

[0269] Figure 1 is a schematic representation of the in-house bioinformatic workflow. ctDNA: circulating tumor DNA; HPV: human papillomavirus; Kraken2.

[0270] Figure 2 is a schematic representation of patient workflow. LA: Locally advanced; SCCHN: squamous cell carcinoma of the head and neck.

[0271] Figure 3 is a swimmer plot of plasma samples and relapse. Each row represents a patient, identified by its Record ID. The length of the line represents duration of followup (normal line means no relapse was observed during the follow-up period; dashed line means a relapse was observed during the follow-up period). The black cross f represents the moment of death. The crossed round represents the moment of relapse. The dots represent plasma evaluation: black dots mean ctDNA was detected in the plasma sample; grey dots mean no ctDNA was detected in the plasma sample.

[0272] Figures 4A-4B is a line graph showing kinetics of plasma variant detection in the two false-positive patients who developed another cancer. Fig. 4A shows patient with lung adenocarcinoma. Fig. 4B shows patient with pancreatic carcinoma. Variants found in the SCCHN are in blue, and variants not found in the SCCHN are in red.

[0273] Figures 5A-5D is a combination of graph showing Kaplan-Meier estimation of progression-free survival (Fig. 5 , Fig. 5C and Fig. 5E) and survival (Fig. 5B, Fig. 5D and Fig. 5F) for MRD-positive and -negative patients in the pre-treatment ctDNA- positive population (n=41) (Fig. 5A and Fig. 5B), in the entire population (n=53) (Fig. 5C and Fig. 5D) and in the HPV-negative pre-treatment ctDNA-positive population (n=31) (Fig. 5E and Fig. 5F). Curves with empty dots = MRD-negative patients; Curves with filled dots = MRD-positive patients.

EXAMPLES

[0274] The present invention is further illustrated by the following examples.

Example 1:

Materials and Methods

Patient cohort, study design, and inclusion criteria

[0275] After obtaining written informed consent, all SCCHN patients treated with standard-of-care protocols were prospectively enrolled into UCL-ONCO-2013 (2013/12FEV/047), a non-interventional biomarker trial collecting whole blood, plasma samples, and sometimes tumor biopsies, before, during, and after therapy (NCT02139020). The clinical characteristics and disease outcomes of each patient were also prospectively recorded in a secured database (REDCap). The study was approved by our ethics committee and conducted in accordance with the Declaration of Helsinki (October 2000).

[0276] Eligibility criteria were: patients (i) enrolled in UCL-ONCO-2013 between 2019 and 2022 with LA- SCCHN (stages III/IVa/IVb pl 6-negative oropharyngeal, larynx, oral cavity, hypopharynx, and unknown primary cancers; and stage I (only N1)/II/III pl6- positive oropharyngeal cancer (OPC) according to the 8th TNM classification), (ii) treated with curative-intent therapy according to standard of care, (iii) for whom pre- and post- treatment plasma samples and whole blood were available, and (iv) with no clinical disease progression from the initiation of curative-intent treatment until collection of the post- treatment plasma sample. The post-treatment plasma sample had to be collected within 12 weeks of the end of the curative treatment. Nasopharyngeal, salivary, and sinonasal cancer were excluded. HPV-driven OPCs were defined as tumors that were positive for HPV by in-situ hybridization (ISH).

Study objective and endpoints

[0277] The main objective was to develop a tumor-agnostic ctDNA assay to detect MRD within 12 weeks of the end of curative-intent treatment in unselected LA SCCHN. The primary endpoint was the rate of progression-free survival (PLS) at two years in the MRD-positive and MRD-negative patients who had ctDNA detected in their pretreatment plasma sample. Secondary endpoints were PLS and overall survival (OS) in MRD-positive and MRD-negative patients.

[0278] MRD was defined as ctDNA detection in the post-treatment sample collected within 12 weeks of the end of the curative treatment. PLS was defined as the time interval between the date of the first day of curative treatment and the date of SCCHN progression, the date of last follow-up or the date of death due to any cause. OS was defined as the time interval between the date of the first day of curative treatment until death due to any cause, or until the date of last follow-up.

Blood and plasma collections

[0279] Plasma samples were harvested from each patient before the start of curative- intent treatment and between one and 12 weeks after the end of curative-intent treatment. When possible, plasma samples were also collected during follow-up, every two to three months until relapse, or for up to three years following a patient’s initial diagnosis. For each patient and time-point, plasma was isolated from whole blood collected in two EDTA 7.5mL tubes after double centrifugation (2000g 10 minutes and 2500g 15 minutes) within 30 minutes at 4°C. Plasma was stored immediately at -80°C until further use.

Cell free DNA (cfDNA) and germline DNA extraction

[0280] cfDNA was extracted from 4mL to 6mL of plasma using the Promega® Maxwell® RSC ccfDNA LV plasma kit (RSC; Promega, Leiden, the Netherlands) and quantified with Qubit® (Thermo Fisher Scientific, Waltham, Massachusetts, USA). Matched germline DNA was extracted for each patient with the Promega® Wizard® Genomic DNA Purification kit (RSC; Promega, Leiden, the Netherlands) from a whole blood sample in EDTA 3.5mL tubes, frozen at -80°C and quantified with Nanodrop® technology (Thermo Fisher Scientific, Waltham, Massachusetts, USA cfDNA targeted next-generation sequencing (NGS) and variant calling

[0281] Plasma samples were analyzed following our in-house workflow, as described in Figure 1. All pre-treatment plasma samples were first analyzed with Kraken2 (v. 2.1.2, database “Standard”, March 2023), a taxonomic sequence classifier that assigns taxonomic labels to DNA sequences. A patient was considered HPV16 positive if at least one read was assigned to the HPV16 genome in the pre-treatment plasma sample, according to Kraken2. Otherwise, the patient was considered HPV16 negative.

[0282] Pre- and post-treatment plasma samples of HPV16 ctDNA-negative patients were analyzed using the gene panel consisting of the most frequently mutated genes in SCCHN according to literature and our own in-house tumor sequencing data. Twenty- four genes plus two additional genes (E6 and E7) from the HPV16 genome were included in the panel (see Table 2). MRD was assessed in each patient through an in-house informatic workflow informed by somatic mutations identified in the corresponding pretreatment plasma sample (see next section for details). [0283] A variant was considered positive in the pre-treatment plasma sample if it was (i) supported by at least three reads, (ii) present in reads aligned in forward and reverse, (iii) passed modified Mutect2 quality filters, and (iv) present in less than one percent of healthy individuals in the public sequence databases (Gnomad). Read alignment for all variants was inspected manually. The list of all criteria can be found in Table 1, and only variants fulfilling all criteria were considered true somatic pre-treatment variants.

Table 1: Filters used for selection of baseline variants

[0284] A pre-treatment sample was considered ctDNA-positive if at least one somatic variant was called. A variant was considered present in the post-treatment plasma sample if it was (i) previously identified in the pre-treatment sample, (ii) called in the BAM file and (iii) detected in three or more reads in the BAM files of the post-treatment plasma sample of the same patient. The ctDNA level was estimated to be “high” if the higher allele frequency per sample was above the median of all variant allele frequencies per timepoint, and “low” if below the median. Libraries for germline DNA

[0285] Libraries for germline DNA (gDNA) from whole blood and cfDNA from plasma were conceived according to the manufacturer’s protocol using MF and EZ Twist Bioscience® (South San Franscisco, California, USA) kits for cfDNA and gDNA, respectively. Targeted NGS was performed after capture using a custom-made in-house panel of 24 genes plus two HPV16 oncogenes E6 and E7) designed by Twist Bioscience®, with an expected mean coverage of 2000x. The 24 genes were selected because they are frequently mutated in HPV-negative SCCHN14 (see Table 2).

Sequencing (‘paired-end’ 2x100 bp reads, Illumina® NovaSeq600™) was carried out by IntegraGen (Evry, France), using unique molecular identifier technology (UMIs).15

[0286] Table 2: Gene panel

[0287] Raw data (.fasta files) were aligned to the reference human genome assembly

(GRCh38) using BWA 0.7.15. UMIs were removed using fgbiotools and Picard. Aligned sequences (.bam files) were processed using GATK 4.2 "BQSR" (for base quality scores recalibration). Germline single-nucleotide variants and small indels were identified using GATK 4.2 "Haplotype Caller" whereas somatic single-nucleotide variants (SNV) and small indels were identified using Mutect218 (both following Broad Institute best practices). The generated variant calls (.vcf) files were annotated, imported and further analyzed on Highlander, the in-house bioinformatics framework of the Genomics and Bioinformatics platform of UCLouvain (PGEN: https://www.deduveinstitute.be/pgen- bioinformatics). Highlander provides extensive variant-annotation, filtering and visualization. Digital droplet PCR (ddPCR) and HPV in-situ hybridization

[0288] In selected cases, mutation detection by ddPCR (Bio-Rad Laboratories, QX200™ ddPCR system, California, USA) was used to confirm the results obtained by NGS and the presence of the HPV 16 genome (Table 3. Data analysis was performed using Quantasoft™ vl.7.4 software according to the manufacturer’s instructions (Bio-rad Laboratories, California, USA). The list of probes and ddPCR results can be found in Table 3.

[0289] Table 3 ddPCR probes and results

HPV In-situ hybridization

HPV-driven tumors were evaluated using in-situ hydridization for high-risk human papilloma viruses (HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58 and 66) (Ventana ISH iView Blue Plus Detection Kit). Staining was performed on a 3pm slide, using a Ventana-Roche GX autostainer (Roche Diagnosis, Basel, Switzerland). The presence of integrated HPV DNA in the nucleus of tumor cells was evaluated by an expert pathologist.

Statistics

[0290] The primary endpoint was the PFS rate at two years in the population of patients who had ctDNA detected in the pre-treatment sample. The hypothesis was that the PFS rate at two years would be < 30% for MRD-positive patients and > 80% for MRD- negative patients. According to the 2- sided Z-test (unpooled variance), a minimum of 32 pre-treatment ctDNA-positive patients with at least two years of follow-up were needed (a= 0.05, power= 0.9).

[0291] PFS and OS were estimated using Kaplan-Meier methods. Patients without any event (progression or death) were censored at the date of last- follow-up. The occurrence of another cancer was not considered to be an event. Treatment differences in PFS and OS were assessed using the log-rank test.

[0292] Cox proportional hazard regression models were used to assess the prognostic value of several parameters on OS and PFS. All analyses were done using SAS (v9.4) and R (v4.2.1).

Results

Patient characteristics

[0293] Between February 2019 and March 2022, 75 LA SCCHN patients were included in our prospective biomarker plasma collection trial. Fifty-three patients met our inclusion criteria (Figure 2) and had pre-treatment and post-treatment plasma samples available. Patient characteristics are described in Table 4 and Table 5 The most frequent tumor site was the oropharynx (n=29, 54.7%). Seventeen (58.6%) patients had pl6-positive oropharyngeal cancer and, of these, nine were HPV-positive by ISH. For one patient, ISH could not be performed due to the insufficiency of tumor cells. Forty-nine (92.4 %) and four (7.6%) patients were treated with curative-intent (chemo)radiation and surgery as primary treatment, respectively. The estimated median follow-up was 31.0 months (range: 2.9-46.5). At two years, the PFS rate was 60.3% (95% CI: 48.0-75.8%) and OS was 78.0% (95% CI: 67.2-90.4%). The median PFS was 34.7 months (1.7-44.0 month) and the median overall survival was not reached. [0294] Table 4. Patient characteristics

ISH: in-situ hybridization; HPV: human papillomavirus; OPC: oropharyngeal cancer; L: larynx cancer; OC: oral cavity cancer; H: hypopharynx cancer; UP: unknown primary; ECOG PS: Eastern Cooperative Oncology Group performance status

Table 5: Patients’ characteristics

[0295] Post-treatment plasma samples were collected between weeks one and 12 (median: 6 weeks) after the end of the curative-intent treatment.

Pre-treatment ctDNA detection rate

[0296] Overall, ctDNA was detected in 41 (77%) out of the 53 pre-treatment plasma samples (Figure 2).

[0297] The pre-treatment plasma samples of the 53 patients were first evaluated for the presence of HPV-16 ctDNA using Kraken2 (Figure 1). The nine patients with HPV- related oropharyngeal cancer diagnosed by ISH, and the patient for whom ISH could not performed, were found to have circulating HPV16 ctDNA in their pre-treatment samples. HPV ctDNA was not detected in the other patients.

[0298] Using the bioinformatic pipeline described in Figure 1, 31 patients (72.1%) out of the 43 HPV 16-negative patients by Kraken2 had at least one tumor variant in their pretreatment plasma samples. Variants were found in the TP53 (67.0%), NOTCH1 (39.5%), EGFR (30.2%), KMT2D (25.5%), and FAT1 (23.2%) genes. Variant allele frequency (VAF) ranged from 0.2 to 20.2% (median: 0.87%) of circulating cfDNA. Details of the variants found in the pre-treatment samples are included in the Table 6. ctDNA positivity in pre-treatment plasma samples was not predictive of OS (P=0.5) nor progression (P=0.66).

Table 6: List and characteristics of variants found at baseline

ctDNA detection within 12 weeks of completion of curative-intent treatment

[0299] Among the 41 pre-treatment ctDNA-positive patients, 17 (41.4%) had ctDNA detected in the paired post-treatment plasma sample collected within 12 weeks of the end of curative treatment. The VAFs ranged from 0.073 to 31.53% (median: 0.29%) in posttreatment plasma samples. Of these 17 HPV 16-negative patients, 14 (82.3%) had disease progression (median time to progression after end of treatment: 10 months (1-30 months months). Twenty-four patients were post-treatment ctDNA-negative, and 20 (83.3%) of these patients did not experience SCCHN relapse. The median time to disease progression for the four post-treatment ctDNA-negative patients who progressed was 14 months (4.8- 27.4 months). The median lead time between ctDNA-positive plasma samples and the time of clinically or radiologically detected disease progression was 4.9 months (0.7- 32.8 months) (Figure 3).

[0300] Overall, for the 41 pre-treatment ctDNA-positive patients, our MRD ctDNA assay had a sensitivity of 77.8%, specificity of 86.9%, positive-predictive value of 82.3%, negative-predictive value of 83.3%, and accuracy of 79.1% for predicting disease progression. In the HPV-negative population with detectable ctDNA in their pretreatment plasma samples (n=31), ctDNA was detected in post-treatment plasma samples in 14 out of 17 patients who eventually relapsed and was not detected in 11 out of 14 patients who did not relapse. Therefore, in the HPV-negative population of this study, our assay showed a sensitivity of 82.4%, specificity of 78.6%, positive predictive value of 82.3%, negative predictive value of 78.6%, and accuracy of 80.7%. Details of our assay’s performance in the different subgroups can be found in Table 7.

Table 7: Tumor-agnostic assay performance in different subgroups

*primary end point population. MRD: minimal residual disease; HPV: Human Papilloma Virus; ctDNA: circulating tumor DNA; CI: confidence interval, NR: not reached

[0301] Four patients were false negatives in that no ctDNA was detected by our assay after treatment despite confirmed disease progression. One of these patients had an HPV OPC, and HPV 16 ctDNA was detected by ddPCR at 0.04 copies/mL in the Kraken2- negative post-treatment sample taken one week after treatment. One HPV-negative patient had no ctDNA detected in the post-treatment plasma sample taken at week one. In this patient, ctDNA was subsequently detected in the post-treatment plasma sample taken 13 weeks after the first. Unfortunately, the two other patients refused further plasma collection. Three patients were false positives in that ctDNA was detected after treatment without SCCHN progression. The sequential plasma samples and the evolution of variant VAFs when available were analyzed. One patient had a variant detected one week after curative-intent chemoradiation. This patient also refused sequential plasma sampling, making it impossible to determine if the variant disappeared thereafter. Interestingly, the two other patients developed non-SCCHN cancers during follow-up (Figure 4). The first developed a lung adenocarcinoma and the second a pancreatic carcinoma at 20 and 12 months of follow-up, respectively. These second cancers were not detectable on 2'-deoxy- 2'-[18F]fluoro-D-glucose positron emission tomography (18FDG-PET) nor on the computed tomography (CT) scans performed at the time of SCCHN diagnosis. For both patients, the SCCHN biopsies obtained at diagnosis were sequenced. The two ctDNA variants found in the pre-treatment sample of the lung adenocarcinoma patient were not found in the SCCHN biopsy and remained present in the plasma after SCCHN treatment (Figure 4). They disappeared from plasma following surgery of the lung adenocarcinoma, suggesting that this was where the variants originated. In the patient with pancreatic adenocarcinoma, seven ctDNA variants were detected in the pretreatment plasma sample. One of these variants was also found in the SCCHN biopsy but was cleared in the post-treatment plasma sample. However, the other six variants were not found in the SCCHN biopsy and were still present in the post-treatment plasma samples (Figure 4), suggesting that they originated from the pancreatic cancer.

The presence ofMRD after curative treatment is associated with lower PFS and OS

[0302] For pre-treatment ctDNA-positive patients, the two-year PFS rate was 23.5% (IC 95%: 10-55%) and 86.6% (IC 95%: 73.4-100%) in the MRD-positive and MRD-negative patients, respectively (p<0.05). Median PFS was 11 months for the MRD-positive patients but was not reached for those who were MRD-negative (p<0.0001) (Figure 5A). Median survival was 28.4 months for the MRD-positive patients but was also not reached for the MRD-negative group (p=0.011) (Figure 5B).

[0303] Sensitivity analyses for PFS and OS in three subgroups were performed: the pretreatment ctDNA-positive HPV-negative population (n=31), the eligible HPV-negative population (n=43), and the eligible population enrolled in this study (n=53). For these two latter analyses, pre-treatment ctDNA-negative patients were included in the MRD- negative group. These additional analyses showed similar results to the ones observed in the primary endpoint population (Figure 5). For the pre-treatment ctDNA-positive HPV- negative population (n=31), the 2-year PFS rate was 23.5.0% (IC 95%: 10.0-55.4%) and 85.7% (IC 95%: 69.2-100%) in the MRD-positive and -negative groups, respectively. Median PFS was 11.0 months [5.8-34.7 months] for the HPV-negative MRD-positive patients but was not reached for those who were MRD-negative (P<0.00033) (Figure 5E). Median survival was 28.4 months [14.3 months-not estimable (NE)] for the MRD- positive patients but was also not reached for the MRD-negative group (P=0.024) (Figure 5F). For the eligible HPV-negative population (n=43), regardless of their pre-treatment ctDNA status, PFS and OS remained worst in the MRD-positive group compared to the MRD-negative group (P=0.00026 and P= 0.068, respectively) (Figures 5G and 5H).

[0304] On the 53 patients that met our clinical inclusion criteria, Cox univariate and multivariate analyses and included several prognostic factors were performed: HPV status, disease stage, primary tumor location, Eastern Cooperative Oncology Group (ECOG) performance status, pre-treatment ctDNA level, type of primary curative-intent treatment and the presence of MRD) (Table 8. In the univariate analysis, having a pretreatment ctDNA level above the median value also significantly correlated with worst PFS (HR=4.3 (1.82-10.39, p<0.001) but not OS (p=0.153). In the multivariate Cox model, MRD positivity remained an independent predictor of PFS (Table 8b) (HR=12.2; 95% CI [2.59-57-50]; p=0.002), and OS (HR=12.22; 95% CI [1.58-94.6]; p=0.017) (Table 8a).

Table 8a : Identification of potential prognostic factors for overall survival

ECOG: Eastern Cooperative Oncology Group performance status; MRD: minimal residual disease; NE: not evaluable. Analysis performed using COX proportional Hazard model. Higher bound for inclusion in multivariate is set to p=0.20

Table 8b: Identification of potential prognostic factors for progression-free survival

Claims

1. A method for predicting a risk of cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

2. A method for detecting a minimal residual disease (MRD) in a subject previously treated for a cancer, comprising the steps of:

(i) isolating cell free DNA (cfDNA) molecules from at least one biological sample obtained from the subject before treatment and from at least one biological sample obtained from the subject after treatment, and optionally amplifying said cfDNA molecules; (ii) sequencing the cfDNA molecules isolated and optionally amplified at step

(i);

3. A method for treating a cancer relapse, in a subject previously treated for said cancer, comprising the steps of:

4. The method according to any one of claim 1 to 3, wherein step (iii) further comprises comparing aligned sequence data obtained from germline DNA.

5. The method according to any one of claim 1 to 4, wherein the step (iii) comprises the following steps:

- measuring a cfDNA variant allele frequency (VAF) in the biological samples.

6. The method according to claim 5, wherein at step (iii), a cfDNA variant is selected when: the cfDNA from said at least one biological sample collected before treatment is characterized in that: (a) at least three reads are detected for said cfDNA; (b) the reads are aligned in forward and reverse; and (c) the cfDNA is present at less than 1% of healthy individuals in public sequence databases; and/or the cfDNA from said at least one biological sample collected after treatment is characterized in that: (a) said cfDNA is also identified in the at least one biological sample collected before treatment; (b) at least three reads are detected for said cfDNA.

7. The method according to any one of claims 1 to 6, wherein a ctDNA molecule from at least one biological sample collected before treatment and a ctDNA molecule from at least one biological sample collected after treatment are considered identical if they share the same nucleic acid sequence.

8. The method according to any one of claims 1 to 7, wherein said at least one biological sample is plasma.

9. The method according to any one of claims 1 to 8, wherein said at least one biological sample collected after treatment is collected at least one week after the end of said treatment, preferably at least 12 weeks after the end of said treatment.

10. The method according to any one of claims 1 to 9, wherein more than one biological sample are collected after said treatment.

11. The method according to any one of claims 1 to 10, wherein the subject was treated for locally advanced squamous cell carcinoma of the head cancer or neck cancer (SCCHN); in particular stage III or stage IV locally advanced SCCHN.

12. The method according to any one of claims 1 to 11, wherein the ctDNA sequence comprises a nucleic acid sequence of any one of the genes selected in the group consisting of: AJUBA, CASP8, CDKN2A, CTCF, EGFR, EPHA2, FAT1, FBXW7, FLG, FOSL2, HIST1H1B, HLA-A, HRAS, HRNR, KMT2D, KTR5, NECAB1, NOTCH1, NSD1, PIK3CA, PSIP1, PTEN, TGFBR2, TP53, E6, E7, or variant and/or a fragment thereof.

13. The method according to any one of claims 1 to 12 wherein the ctDNA sequence comprises a nucleic acid sequence selected from E6 or E7.

14. The method according to any one of claims 1 to 12, wherein said SCCHN is HPV- positive SCCHN, preferably HPV-16 positive SCCHN.

15. A computer program comprising instructions which, when the program is executed by a computer, causes the computer to execute the steps of the method according to any one of claims 1 to 14.

16. A computer-readable medium having stored thereon the computer program of claim